Although the Left Hemisphere (LH) appears dominant for processing language, a great deal of evidence indicates that the right Hemisphere (RH) also processes language. Predictions from a new theory of RH language processing will be tested in a five-year program of experiments. This theory proposes that when processing a work the RH weakly activates a large "field" of related semantic representations, including peripherally related information. However, the activate is too diffuse to select information for further processing. This is termed coarse semantic coding. In contrast, the LH strongly activates a narrow semantic field of closely related information, enabling it to select representations for further processing or for consciousness. This is termed fine semantic coding. This theory predicts that RH language processing will be maximized in experiment tasks that a) utilize peripheral information; b) utilize overlap from multiple inputs; and/or c) not require selection of a single unique response. The proposed methods will equate performance of controls and patients with RH or LH lesions, and of normal subjects processing linguistic stimuli directed to each hemisphere, revealing qualitatively different patterns of performance. In the proposed project, 24 experiments will test the following general hypotheses: 1) That drawing coherence inferences depends on the RH homologue to Wernicke's area, because this area is important for recognizing distant relations between words. 2) That the semantic activation of peripherally related information depends on this same ares, and as a result the RH is more sensitive to particular types relations between words. 3) That the ease with which people can read words from each part of speech is modulated by differences in the semantic information conveyed by these words (as seen in alexic patients), and that the RH is more sensitive than the LH to these intrinsic word differences. 4) That subthreshold semantic activation in the RH is too weak to elicit explicit identification or awareness, but can influence binary decisions about broad semantic categories ("Is it a fruit") . 5) That the LH includes a layer of processing absent in the RH. The RH coarse semantic coding theory is pitted against two alternatives: That attentional mechanisms focus activation more sharply in the LH than in the RH; Or, that homotopic interhemispheric inhibition shapes the semantic activation of both hemispheres. According to the proposed theory, complete discourse processing requires coordination of both RH coarse semantic coding and LH fine semantic coding. The proposed project is expected to enhance knowledge of the specific brain regions and mechanisms underlying language processes. Such knowledge will inform future models and computer simulations of language processing, neuroscientific examination of anatomical asymmetries, and clinical recognition of the language impairments of patients with RH lesions and the residual language abilities of alexic patients with LH lesions.